Where to Start?
There is a lot of information available out there. The hard part seems to be figuring out where to start. For example, see this recent thread on HackerNews that generated quite a lot of nerd.
- neuralnetworksanddeeplearning.com - Online Book. Introductory. Seems well-written and well-recieved.
- TensorFlow & Deep Learning without a PhD - Short 5-hour class, on YouTube
- The Naked Tensor
- CS20si - TensorFlow for Deep Learning
- Practical Deep Learning For Coders, Part 1 - Free online course, looks pretty good.
- TensorFlow Tutorials - Definitive guide to the most popular deep learning tool.
- TensorFlow MNIST Tutorial - Someone's personal GitHub TensorFlow intro talk, with code.
- TensorFlow MNIST for ML Beginners - Official TensorFlow tutorial
- DeepLearning.net - Theano / Python introduction to deep learning tools and broad concepts
- Stanford Deep Learning Tutorial - 2014. Hacker News nerd rage because it is so old. But it looks like it covers the basics really well to me.
- scikit Learn - Examples - Seems fun and hands-on for an important ML tool. No theory at all.
- Stanford CS231n - Convolutional Neural Networks for Visual Recognition. May not be beginner friendly? Seems very well recieved. Likely to be great.
- Stanford CS224d - Deep Learning for Natural Language Processing. In Python. Not for beginners. Looks super fun.
- Colah Blog - Lots of nice original posts and papers. Some of it not beginner friendly, but seems well-written.
- Coursera ML Class - Introductory. Probably great. Graded. Need 2 Enroll.
- R-Bloggers ML Class - Statistical ML Class. 15 Hours of videos. It looks like a great introduction. But it is in R.
- 40 Data Science Techniques - Not ML specific, but still a good jumping-off point.
- Data Science Toolset - Blog post with good list at the bottom:
- TensorFlow Playground - A toy, not a resource. But you can play with a little neural network, tweaking various knobs. And if you don't know what a knob does, you know you need to go look that up.
- Stanford Online ML Degree - Not a Resource. Four online classes required. Looks cool.
- ML from Scratch - Just a fun collection of ML ideas sketched out in Python to play with.
- NEAT - A YouTube video with links to information on Neural Evolution using Augmented Topologies (also see paper)
- spaCy - NLP in Python
Honestly, I have not found a lot of really great books on these topics. There are some good theory text books, but they are all a bit out of date because of how long it takes to publish a book. And there are not a lot of practical books on Amazon to learn to use specific tools like SciKit Learn or TensorFlow.